Transcript Multistage Sample

DATA COLLECTION METHODS
SAMPLING
1
2
Class Objective
After this class, you will be able to
- Use Non-probability Sampling Method
- Identify difficulties and disasters in
sampling
Multistage Sampling
Using a combination of the sampling methods,
at various stages.
Example:
• Stratify the population by region of the country.
• For each region, stratify by urban, suburban, and
rural and take a random sample of communities
within those strata.
• Divide the selected communities into city blocks as
clusters, and sample some blocks.
• Everyone on the block or within the fixed area may
then be sampled.
Copyright ©2011 Brooks/Cole, Cengage Learning
3
The Current
Population Survey
Unemployment rates in U.S. estimated each month based on
interviews with sampled households as part of the Current
Population Survey by the Bureau of Labor Statistics.
Multistage Sample:
• Stage 1: country divided into ~2000 PSUs (primary sampling
units) based on population and size; within each state,
strata of PSUs created based on unemployment rates;
one PSU is randomly selected from each of the strata.
• Stage 2: USUs (ultimate sampling units) randomly selected
from each PSU. All households in the selected USUs visited.
• Stage 3: If number of households in Stage 2 still too many, a
systematic sample of the households is obtained.
Copyright ©2011 Brooks/Cole, Cengage Learning
4
Not Using a Probability Sample
Responses from a self-selected group, convenience
sample or haphazard sample rarely representative
of any larger group.
A Meaningless Poll
“Do you support the President’s economic plan?”
Results from TV quickie poll and proper study:
Those dissatisfied more likely to respond to TV
poll and it did not give the “not sure” option.
Copyright ©2011 Brooks/Cole, Cengage Learning
5
Case Study
The Infamous Literary
Digest Poll of 1936
Election of 1936: Democratic incumbent
Franklin D. Roosevelt and Republican Alf Landon
Literary Digest Poll:
• Sent questionnaires to 10 million people from magazine
subscriber lists, phone directories, car owners, who
were more likely wealthy and unhappy with Roosevelt.
• Only 2.3 million responses for 23% response rate.
Those with strong feelings, the Landon supporters
wanting a change, were more likely to respond.
• (Incorrectly) Predicted a 3-to-2 victory for Landon.
Copyright ©2011 Brooks/Cole, Cengage Learning
6
Case Study
The Infamous Literary
Digest Poll of 1936
Election of 1936: Democratic incumbent
Franklin D. Roosevelt and Republican Alf Landon
Gallup Poll:
• George Gallup just founded the American Institute of
Public Opinion in 1935.
• Surveyed a random sample of 50,000 people from list
of registered voters. Also took a random sample of 3000
people from the Digest lists.
• (Correctly) Predicted Roosevelt the winner. Also predicted
the (wrong) results of the Literary Digest poll within 1%.
Copyright ©2011 Brooks/Cole, Cengage Learning
7
8
Class Task
• What could go wrong in sampling?
• You have 5 minutes
Difficulties and
Disasters in Sampling
Some problems occur even when a
sampling plan has been well designed.
• Using wrong sampling frame
• Not reaching individuals selected
• Nonresponse or nonparticipation
• Self-selected sample
• Convenience/Haphazard sample
Copyright ©2011 Brooks/Cole, Cengage Learning
9
Using the Wrong Sampling Frame
The sampling frame is the list of units from which the
sample is selected. This list may or may not be the same
as the list of all units in the desired “target” population.
Example: using telephone
directory to survey general
population excludes those
who move often, those with
unlisted home numbers, and
those who cannot afford a
telephone. Solution: use
random-digit dialing.
Copyright ©2011 Brooks/Cole, Cengage Learning
10
Not Reaching the Individuals Selected
Even with a properly selected sample,
researchers might not reach the desired units.
•
•
•
•
Telephone surveys tend to reach more women.
Some people are rarely home.
Others screen calls or may refuse to answer.
Quickie polls: almost impossible to get a
random sample in one night.
Copyright ©2011 Brooks/Cole, Cengage Learning
11
Nonresponse, Volunteer Response,
and Nonparticipation
Even with a properly selected sample, individuals
cannot be contacted or refuse to participate.
• Volunteer responses not likely to represent
the entire sample, and leads to nonresponse or
nonparticipation bias.
• Response rates should be reported in summaries.
Copyright ©2011 Brooks/Cole, Cengage Learning
12
Which Scientists
Trashed the Public?
“82% (of scientists) trashed the media, agreeing with the
statement ‘The media do not understand statistics well
enough to explain new findings.’ ” Science (Mervis, 1998)
Science Poll
• 1400 professionals (in science and in journalism).
• Only 34% response rate among scientists.
• Typical respondent was white, male physical scientist
over age of 50 doing basic research.
• Respondents represent a narrow subset of scientists
 inappropriate to generalize to all scientists.
Copyright ©2011 Brooks/Cole, Cengage Learning
13
Homework
• Assignment:
– Chapter 5 Exercise 5.57 – 5.60
• Reading:
– Chapter 5 – p. 166 - 170